DE3809791A1 - METHOD AND DEVICE FOR CARRYING OUT MAGNETIC RESON EXAMS WITH A LIMITED VOLUME - Google Patents

Description

The invention relates to nuclear magnetic resonance imaging (MRI) and especially on methods for spatial Limit the volume of interest for the MRI. As in applications of the applicant that are of interest in this connection is filed on Israeli application 78,240 on March 24, 1986, to US application 6 52 462, filed on September 20, 1984 and to Israeli application 76,398, filed on September 13, 1985.

The technique of chemical shift spectroscopy with Nuclear magnetic resonance (NMR) is in relatively long time Application. In 1973, P. C. Lauterbur in one in Nature (London) 242, pages 89/90 published article on the use of field gradients to determine the Source location of signals with free induction decay (FID) pointed out, which obtained in nuclear magnetic resonance experiments have been. Knowing the source of the FID signals enables that the data acquired by magnetic resonance is used to reconstruct interior images of the object, which is brought into a strong magnetic field.

It has long been known that when atomic nuclei, the possess no magnetic moments, in a strong static Magnetic field are brought, these nuclei (spine) one Precession movement around the axis of the field with the Larmor frequency perform, which is given by the equation:

f = γ · B ₀ / 2 π

in which

γ = a gyromagnetic ratio that is constant for each NMR isotope that has a usable magnetic moment, B ₀ = the strength of the magnetic field, π = the constant 3.1416. . .

As is known, a relatively strong static magnetic field with a given direction, which is aligned with the Z axis of the Cartesian coordinate system, is used in magnetic resonance imaging (MRI). The strong static magnetic field causes the cores or "spine" of certain elements, e.g. B. of hydrogen, align with the field. As a result, RF pulses with a sufficiently high amplitude and / or long duration are applied to disrupt or tilt the aligned "spine". The frequency of rotation of the RF precision and the frequency of the RF pulses is the aforementioned Larmor frequency.

After the RF pulse has ended, the rotated spins tend to realign with the static magnetic field. The precession of the transverse component is generated in the magnetic field RF signals that also have a Larmor frequency. These Signals are signals with free induction decay (FID) known. It is these signals that are received at a Information regarding the spin density of the element achieve, whose spine rotated by the RF pulse have been. The spin density information becomes an image used.

There are a number of different methods used to Achievement of the FID signals can be applied. One of these Methods, and perhaps the best known at the moment, is Spin echo method, which is widely known and therefore not is discussed in more detail.

The scientists are always in the picture display strives on the one hand to improve the spatial resolution and on the other hand reduce the amount of time required is to represent the picture. This is it to contradict each other because of a reduction the duration in general a decrease in resolution entails and the signal-to-noise ratio disadvantageous influenced. A method of reducing the amount of time maintaining the same resolution at the same time  and / or the same signal-to-noise ratio, or a method of increasing resolution while at the same time Therefore, image display during the same period is asked. In magnetic resonance imaging, one brings No increase in the duration of the acquisition of an image Danger to the patient because no dangerous radiation is used; because patient comfort and throughput speed important considerations are both the Image quality as well as the economy of the system are clinicians and imaging specialists always interested in capturing images reduce the time required. In some cases The time saved can be used to take multiple pictures of the collect the same disc and then the different images to average the signal-to-noise ratio to improve.

Another goal of the imaging specialists is to zoom during the capture stage. In other words this means that during the imaging process, if a certain part of the body a symptom of interest shows, is often desired to zoom in on this symptom and thus excluding the consideration of this symptom focus on other data. So far this has been common with MRI systems through a computer step after acquisition of the data, especially if the image is within within a certain time frame. By such manipulation of the data cannot Improvement in spatial resolution can be achieved. The third of the patent applications mentioned above gives a Procedure for zooming in while collecting data. A Such zooming can change the resolution of the part of the image increase, focused on in a natural way has been.

A problem with zooming while collecting data occurs, is that "aliasing" artifacts that caused by under-examination, can be generated if the number of coding cycles is not related  with a proportional increase in the total acquisition time is increased. The relationships between the field of view the resolution and data acquisition time are as follows:

The size of the volumetric acquisition matrix is

n _x · n _y · n _z

where n _x , n _y and n _z denote the sizes of the matrix along the X , Y and Z axes.

The volume of a voxel is

V = l _x · l _y · l _z ,

where l _x , l _y and l _{z are} the dimension along the X , Y and Z axes.

The field of view FOV is FOV = li · ni , where i = x, y, z .

The resolution L for voxel n is L = n _i / FOV .

The data acquisition period Ta is Ta = TR * n _x * n _y (assuming phase coding along the X and Y axes), where TR is the repetition period.

Limiting the field of view increases the resolution with a fixed registration matrix. In a similar way reduces the limitation of the field of view with one fixed resolution the acquisition duration.

The location of the volume of interest is in critical critical to medical, diagnostic Use cases of magnetic resonance spectroscopy (MRS) and is useful for magnetic resonance imaging (MRI). The So far, a cubic volume has been selected by a Variety of techniques and systems achieved. For example the application of RF pulse trains involves three successive ones  special RF pulses, each of which is subsequently other of the three orthogonal gradients to select one desired cubic volume can be used.

The use of pulse trains, e.g. B. 90 °, 180 ° and 180 ° was entitled by R.E. Gordon et al in a report "Volume Selection for High Resolution NMR Studies" in "Processes of the SMRM Third Annual Meeting", 1984, pages 272 ff explained.

Using a pulse train for spatial localization a composite pulse, e.g. B. selective 45 °, nonselective 90 ° and selective 45 °, the composite Impulse was given three times, each time with one of the three orthogonal gradients was included in an essay entitled "A Selective Volume Method for Performing Localized NMR Spectroscopy "by W. P. Aue et al in" Journal of Magnetic Resonance ", volume 56, pages 350 ff. The process according to this article is the subject of US-PS 44 80 228.

A pulse train for spatial localization in spectroscopy using combinations of three selective 180 ° pulses and a non-selective 90 ° pulse is in an essay entitled "Image-Selected in Vivo Spectroscopy (ISIS). A New Technique for Satially Selective NMR Spectroscopy "by R. J. Ordidge et al in" Journal of Magnetic Resonance ", volume 66, pages 283-294 (1986).

Another pulse sequence for a spatial localization of the Spectroscopy is in an essay entitled "H MR Spatially Resolved Spectroscopy of Human Tissues in Situ "by P.R. Luyten et al, published in "Magnetic Resonance Imaging ", Volume 4, pages 237-239 (1986).

Another pulse sequence for the volume selection in the Magnetic resonance spectroscopy is the subject of an article entitled "Spatial and Chemical - Shift-Encoded Excitation.  SPACE, a New Technique for Volume-Selected NMR Spectroscopy " by D. M. Doddrell et al, published in "Journal of Magnetic Resonance, Vol. 68, pages 367-372 (1986).

The known pulse train processes with selective 90 ° -180 ° -180 ° and the non-selective 90 ° -180 ° and selective 90 ° for the spatial localization of the received NMR signals result in signals which strongly depend on the T ₂ recovery times of the spins. This dependence on the T ₂ recovery times makes it difficult to display signals with short T ₂ recovery times.

The method proposed by Aue et al. H. the pulse train with selective 45 °, non-selective 90 ° and selective 45 °, enables a very high RF energy and has that Disadvantage of precision influences outside the resonance (see essay by Doddrell et al), which during the compound impulse transmission occur. These effects adversely affect the signal-to-noise ratio (SNR).

The method proposed by Ordidge et al is sensitive to subtraction noise and requires extremely precise magnetic field stability to ensure the exact cancellation of signals obtained from spins that are not in the volume of interest (VOI) . Instrumentation problems may also arise that counteract the effective display of weak signals.

The Doddrell et al method requires a relative one high power requirement and has a relatively high "Subtraction noise".

The method according to the second patent application mentioned above has a low power requirement and a T ₂ dependence, but uses the signals from simulated echoes compared to signals from a full echo.

The aim of the invention is therefore to create a method and a system (or a device) for carrying out magnetic resonance examinations with limited volumes, which uses a complete echo sequence in order to obtain data from selected volumes for use in spectroscopy or in imaging, which are not highly dependent on T ₂, which effectively limit RF pulse power requirements and sensitivity to subtraction noise, and which also effectively limit the acquisition of signals to the selected volume of interest. Another object of the invention is to create MR image representations from limited volumes with the aid of spatially localized signals using selected pre-inversion pulses as part of the method of volume selection.

According to the invention, a method for performing MR examinations with limited volume is proposed, which is characterized in that
Spine are aligned in a test specimen by positioning the test specimen in a homogeneous static magnetic field in order to obtain data derived from NMR from the test specimen,
a first basic scanning sequence is carried out, in which the test object is irradiated with a first RF pulse in the presence of a first magnetic gradient so that the aligned spins are inverted in a flat disk of the test object and
a modified spin echo sequence is applied in order to achieve a selected strip with a partial strip, the spines being 180 ° out of phase with the spins in the remaining part of the selected strip, and the partial strip being part of the flat disk,
a second basic scan sequence is applied, by which the modified spin echo sequence is applied to obtain the stripe without the partial stripe, and
the strips of the first and second modified spin echo sequences are combined to obtain only the partial strip.

With the invention unwanted FID signals and Echo signals eliminated or suppressed.

In a further embodiment of the invention, selection gradients are used which extend beyond the first RF pulse and which start before the second RF pulse in order to disturb coherence, or they become spoiler gradients in the time interval between the second and third RF pulses given up. These gradients can be applied along the X and / or Y and / or Z axis.

Furthermore, the invention provides for the abandonment of a separate one Gradient pulse along the viewing axis after the Application of the second RF pulse for the phase shift in front. A coding gradient can also be given if a two-dimensional projection of the three-dimensional Disk to be imaged; continue two Coding gradients along two axes are given, if a three-dimensional projection of the three-dimensional Disk to be mapped.

Another feature of the invention provides the abandonment of the Coding gradient or the gradient and / or the abandonment of the phase shift gradient along the viewing axis between the first and second RF pulses or between the third RF pulse or the viewing gradient pulse.

It is also proposed with the present invention that both the position and the size of the volume can be easily selected in that a corresponding frequency and / or bandwidth of the three RF pulses along with controlling the strength of the selection gradients is selected. The choice is made according to the following equations:

Slice position x = 2 fΔ f ₀ / Gx and slice width x = 2 πΔ F / Gx

in which

x = the position along the X axis (can also be the Y or Z axis), f ₀ = an offset frequency (which is added to the Larmor frequency), F = bandwidth of the RF pulse.

Finally, the present invention proposes that the selective inversion process is the simultaneous Collection of data from multiple localized volumes under Use of procedures according to the Israeli patent application 76 009 of 2 August 1985.

The invention will now be read in conjunction with the drawing explained using an exemplary embodiment. It shows:

Fig. 1 is a block diagram of a typical NMR data acquisition system,

FIG. 2 shows two basic RF pulse sampling sequences which contain the complete sampling sequence which is used to obtain the image data from a limited volume according to the invention, FIG.

Fig. 3 is an image sequence using a selected volume of interest in a selected strip in a selected slice of a specimen using the two Grundabtastfolgen according to Fig. 2, and

Fig. 4 graphically shows the FID signals and echoes that are normally generated by three pulse trains.

Fig. 1 shows at 11 a block diagram of an in-vivo magnetic resonance data acquisition system which has a magnet 12 for generating a high static magnetic field. The dimensions of the magnet are such that it has an opening that fits into a patient 14 . The magnetic field is generated by a magnetic field generator, which is designated as block Ho with 13 . The NF magnetic pulses are generated by the HF generator 16 . The pulses are shaped by a modulator 17 . The shape of the pulses can have a Gaussian or Sinc shape as determined by the modulator frequency generator 18 . The shaped pulses act in the presence of a selective gradient in such a way that the area of the test specimen which is acted upon by the RF pulse is limited.

The modulated RF pulse signal is transmitted to body coils in the magnet, which are not shown. The coils can also be surface coils or head coils. The RF pulses are in the presence of gradients, for. B. X -, Y - or Z gradients, which are generated by gradient generators 19, 21 or 22 . The first RF pulse P ₁ ( Fig. 2, 4) is an inverting pulse, ie a 180 ° pulse, which is transmitted by the transmitter 20 in the presence of a gradient pulse GY ( Fig. 2) to the aligned spine to invert in a selected flat portion I ( Fig. 3) of the volume of patient 14 to be examined. In a preferred embodiment of the invention, the first RF pulse is a composite inverting pulse. The composite pulses are less sensitive to offset effects.

At some point, e.g. B. t ₁ after giving up the first RF pulse P ₁, the second RF pulse P ₂, a 90 ° pulse, is given up. The second RF pulse is also preferably shaped. It is given in the presence of a magnetic field gradient that is perpendicular to the gradient that was given during the application of the first pulse P ₁. For example, the gradient Gx that is generated by the gradient generator 19 is canceled. If the magnetic gradient is present in the Y direction, the 90 ° pulse tilts the spine in a disk II by 90 °. The disk II has a strip III, which is part of the disk I, in which the spins are now 180 ° out of phase with the other spins in the disk II. Thus, the first RF pulses P ₁ the aligned spine in the disc I by 180 °. The second RF pulse P ₂ tilts the spine in disc II by 90 °. Since the spines in strip III were previously inverted when they are also tilted by 90 °, they remain inverted with respect to the previously non-inverted spins.

At some point, e.g. B. t ₂ ( Fig. 4) after giving up the second RF pulse P ₂, a third RF pulse P ₃ is given. The pulse P ₃ is a second 180 ° pulse, which is preferably also shaped. The third RF pulse P ₃ is given in the presence of the gradient pulse Gx , which is generated by the gradient generator 19 perpendicular to the other two gradient pulses. The third RF pulse excites the spine and produces echoes from a streak labeled IV. The excitation of the spine in strip IV includes the spine of sub-strip V. The sub-strip V is the volume of interest VOI . It is the strip that is formed by the intersection of strips III and IV. Thus the connection of strips III and IV defines the volume V.

The received signals are indicated by receiver coils in the magnet (not shown). The receiver coils and the transmitter coils can be the same. The received signal is received by the receiver 23 and demodulated by the demodulator 24 . The demodulated signal is amplified and processed in the analog / digital processing unit 26 in order to obtain frequency data, which is denoted by 27 . The frequency data are converted into time data 29 by a Fourier transform operator 28 . The time data is processed in the image processor 31 to generate the image for the visual display unit 32 . The entire process is monitored and controlled by processor 33 .

The pulse sequences are shown in Fig. 2. The first shaped RF pulse P ₁ is given in the presence of a Y gradient 36 . This selective pulse causes a pre-inversion of the spine in the selected slice I ( Fig. 3a). The disk I lies in or parallel to the XY plane, as shown by the coordinate system shown in FIG. 3 (g).

After a period of time t ₁ the second shaped RF pulse P ₂ is given up. In a preferred embodiment, it has a sufficient length of time to rotate the HF spine in a selected test specimen by 90 °. The pulse P ₂ is given in the presence of a Z gradient 37 , so that only the spins in the disk II are affected in the patient. The disc II lies in or parallel to the XY plane.

At a time t ₂ after tasks of the pulse P ₂, a third, shaped RF pulse P ₃ is given in the presence of an X gradient 38 to selectively excite the spine in the strip IV, which is in the YZ plane or parallel to this level. In a preferred embodiment, a "spoiler" gradient pulse 38 'is used between the RF pulses P 1 and P 2.

Applying the three RF pulses and the three gradients in the manner described above selects the strip IV. The size and location of the volume is controlled by the processor 33 , which controls the characteristic properties of the gradients and the amplitude, duration and frequency of the pulses. At a time t ₂ after the pulse P ₃ is given, the echo signal E ₂₃ is received in the presence of a viewing gradient 39 .

In a preferred embodiment, the second gradient pulse 37 according to FIG. 2 has a negative part 38 , which is applied after the associated RF pulse has ended. The negative part 38 extends over approximately half the area of the positive part of the gradient pulse 37 and is used to refocus the spine.

The third gradient pulse also has a negative part 39 , which is applied after the third RF pulse has ended. The negative part extends a little over half the range of the positive part of the same pulse.

The spoiler gradient pulse 38 ' and / or the phase shift pulses 39, 40 which are assigned to the gradient 36 , and the phase shift pulses 41, 42 which are assigned to the gradient pulse 37 are in the time intervals between the pulses P ₁ and P ₂ and immediately shown before or after the third RF pulse P ₃ shown to disrupt coherence.

A phase coding gradient is shown in Fig. 2 as given between the pulses P ₂ and P ₃ of the first sequence along the Y axis. A single phase encoding gradient, e.g. B. with 43 shown, results in a two-dimensional projection image of the selected volume. The application of a second phase coding element would result in a three-dimensional projection image representation of the selected volume.

The second basic scan sequence of Fig. 2 is the modified spin echo sequence used in the first basic scan sequence, except that the selective pre-inversion pulse (RF) is not used. The second scanning sequence according to FIG. 2 thus begins with a Gy gradient pulse 51 . The pulse 51 is the same as the gradient pulse 37 of the first scan sequence. The gradient pulse 41 is used even in the absence of the inversion pulse P ₁, because all operations are performed to ensure that the strips obtained by the samples, are identical (except for the inversion of the pulse P ₁) to thereby subtraction errors so to keep it as low as possible. For example, eddy currents are generated by gradients; therefore, if the first basic sequence has a gradient that cannot be found in the second basis, eddy current effects will occur that will not be resolved when the results of the first and second basic frequencies are combined.

Then an RF pulse P ₂ is transmitted in the presence of the Gz gradient pulse 52 . The pulse 52 has a negative part 53 similar to the negative part 38 of the gradient pulse 37 of the first scan sequence.

As in the first scan sequence, phase encoding pulses 54 are applied . A spoiler gradient pulse 56 is applied simultaneously. Then the RF pulse P _{3 'is given} during the application of the X gradient pulse 57 . The viewing gradient pulse 58 is applied during the reception of the echo. Other spoiler and phase shift pulses 56, 59, 61 and 62, 63 are given between pulses P _{1 '} and P _2' and on both sides of the RF pulse P _{3 '} . In addition, the gradient impulses mimic the first sequence in their amplitudes and in their time behavior. So the pulse 52 extends before the pulse P _{2 '} and the pulse 57 on both sides of the pulse P _3' . The pulse 51 extends behind the position of the missing pre-inversion pulse P _{1 '} .

The effects of the second scan sequence are shown in Figures 3d-3f. Figure 3d shows the volume of the specimen in the high static field, with the spins aligned and without pre-invasion. Fig. 3c shows the disc II, which responds as a function of the selective pulses P _{2 '} without pre-inverted strips. Fig. 3d shows the strip IV ', which is obtained in response to the selective pulse P _3' . There is no streak V '. By combining the strips IV and IV ', z. B. by complex subtractions, there is only twice the signals of the strip V (the desired volume of interest).

A method for carrying out the "multi-volume operation" consists of running the full sequence several times during the To give up the repetition time interval. Every repetition chooses a different volume somewhere within the contemplating organ. The only limitation to avoid Signal saturation is that there are no two volumes share a common level.

An effective way to achieve a "multi-volume sensing operation" is to apply n pre-inversion pulses before P ₂ , each with a different offset frequency, to invert the spine of n disks within the device under test. A corresponding choice of the offset frequencies P ₂ and P ₃ determines the end positions of the n volumes of interest. To avoid signal saturation, care must be taken to ensure that no two selected volumes share a common plane.

The system can also be used for a three-dimensional image display used by a second phase encoding gradient pulse along another axis.

The requirement for a minimum of two acquisitions can be considered a disadvantage if the system described to locate the image representation is used. However, since in practice the field of view can be small in relation to the size of the volume of interest, a signal message may be necessary to improve the signal-to-noise ratio. Localized in vivo spectroscopy usually also requires signal averaging. For a maximum signal-to-noise ratio from the volume of interest, it is necessary that t ₁ « T ₁ and t ₂« T ₂, where t ₁ and t ₂ are the time intervals between P ₁ and P ₂ and between P ₂ and P ₃ are. It should be noted, however, that if these conditions are not met, the only result will be a certain loss of sensitivity, but not a disturbance of the signal from the volume of interest by signals from outside the volume of interest. In spectroscopy applications, time-dependent magnetic fields, which are caused by eddy currents which are induced in the magnetic bore by the switching gradients, can bring about a disturbance in the spectral resolution. This influence can be brought to a minimum by extending t ₂ in order to introduce a greater delay between the generation and the signal display (but with increasing dependence of the signal strength on the T ₂ relaxation).

Under conditions of non-ideal RF pulses or inhomogeneous RF fields, such as are found, for example, in whole-body MRI, especially when surface coils are used, care must be taken to ensure that unwanted signals that may have been generated are suppressed. Such erroneous signals can include the FID signals that on the RF pulses (FID ₁, FID ₂, FID ₃), the two pulse echoes E ₁₂ and E ₁₃, and the regular (E ₁₂₃) and stimulated (STE ) echoes which have been produced by the three pulses P _1, P ₂ and P _3, follow.

As described, an elimination of the undesired signals can be achieved with the aid of "spoiler" gradients, and a phase cycle circuit can also be used. So z. B. the extension of the selection gradient from P ₁ to P ₂ or the P ₂ to P ₂ bring any coherence generated by P ₁ out of phase. This also serves to prevent the formation of E ₁₂, E ₁₃ and E ₁₂₃. A symmetrical expansion of the selection gradient from P ₃ to P ₃ as well as additional false gradients along the other axes, which are set before or after P ₃ , eliminates or suppresses DID ₂, FID ₃ and the stimulated echo. An alternating inversion of the phase of P ₁ and summation of the detected signals (after P ₃ ) eliminates E ₁₂, E ₁₃, E ₁₂₃ and the stimulated echo. An alternating inversion of phase P ₂ and alternating addition / subtraction of the displayed signals eliminates E ₁₂, E ₁₃, E ₁₂₃ and FID ₃. A four stage phase cycle with phase inversions of both P ₁ and P ₂ essentially eliminates all unwanted signals.

This describes a system for recording imaging data limited in volume, in which the dependence on the T ₂ recovery time is reduced. The RF energy that is applied during the designated sequence is significantly less than that required using the normal spin echo sequence. The technique described above results in a particularly expedient spatial localization for MRI and a dominant T ₁ contrast in the image display compared to the dominant T ₂ contrast of the regular spin echo image displays.

Claims (24)

1. A method for performing magnetic resonance (MR) examinations with limited volume, characterized in that

• a) aligning the spine in a body by placing the body in a strong magnetic field,
• b) a first basic scan sequence is used which comprises
  • b1) irradiating the body with a first RF pulse in the presence of a first magnetic gradient to invert the aligned spins in a first selected volume of the body,
  • b2) irradiating the body with a second RF pulse in the presence of a second magnetic gradient perpendicular to the first magnetic gradient in order to tilt the aligned spins into a transverse plane in a selected disk of the selected volume,
  • b3) irradiating the body with a third RF pulse in the presence of a third magnetic gradient perpendicular to the first and second magnetic gradients to obtain signals from a selected strip of the selected slice,
• c) a second basic scan sequence is used which comprises the steps of the first basic scan sequence without the first RF pulse, and
• d) the signals of the first and second basic scan sequences are combined to obtain data from a selected volume of interest.

2. The method according to claim 1, characterized in that the first and third RF pulses are 180 ° pulses, the second RF pulse is a 90 ° pulse, and the ones obtained Signals are echo signals.   3. The method according to claim 1, characterized in that the first pulse is a composite inverting Impulse is. 4. The method according to claim 1, characterized in that the first and second basic scan sequences of phase encoding gradient pulses to achieve a to achieve two-dimensional image. 5. The method according to claim 4, characterized in that the phase encoding gradients along the axis of the first magnetic gradients. 6. The method according to claim 1, characterized in that the first and second basic scan sequences two each Sets of encoding gradient pulses include the To achieve a three-dimensional image. 7. The method according to claim 6, characterized in that a first set of the two sets of coding gradient pulses along the axis of the first magnetic gradient pulse runs, and that the second sentence of the two Sets of coding gradient pulses along the axis of the second magnetic gradient pulse. 8. The method according to claim 1, characterized in that in the basic scan sequences of spoiler and phase shift pulses in the basic scan sequences. 9. The method according to claim 1, characterized in that Spoiler gradient pulses between the first and second RF pulses are given up. 10. The method according to claim 1, characterized in that to eliminate unwanted signals asymmetrical Gradient pulses are used.   11. The method according to claim 1, characterized in that n first RF pulses are applied before the application of the second RF pulse, each of the n first pulses having a different offset frequency for obtaining signals from a plurality of selected volumes of interest . 12. The method according to claim 1, characterized in that only the second basic scan sequence is used to To obtain signals from a selected strip. 13. Device for performing magnetic resonance (MR) examinations with a limited volume, characterized by

• a) a device for aligning the spine in a body or test specimen by placing the body in a strong static magnetic field,
• b) a device for giving up a first basic scan sequence
  • b1) a device for irradiating the body with a first RF pulse in the presence of a first magnetic gradient, such that the aligned spins are inverted in a first selected volume of the body,
  • b2) a device for irradiating the body with a second RF pulse in the presence of a second magnetic gradient perpendicular to the first magnetic gradient, so that the aligned spins are tilted in a transverse plane in a selected disk of the selected volume,
  • b3) a device for irradiating the body with a third RF pulse in the presence of a third magnetic gradient perpendicular to the first and second magnetic gradients in order to obtain signals from a selected strip of the selected disk,
• c) a device for applying a second basic scanning sequence which contains all the pulses of the first basic scanning sequence without the first RF pulse, and
• d) an image processing device which combines the signals of the first and second basic scanning sequences with one another in order to obtain data from a selected volume of interest.

14. Device according to claim 13, characterized in that the first and third RF pulses are 180 ° pulses and the second RF pulse is a 90 ° pulse, and that the obtained signals are echo signals. 15. The device according to claim 13, characterized in that the pulse is a composite inverting pulse is. 16. Device according to claim 13, characterized by a Device for applying phase encoding gradient pulses during the first and second basic scan sequences, to get a two-dimensional image. 17. The device according to claim 16, characterized by a Device for applying the phase coding gradients along the axis of the first magnetic gradient. 18. Device according to claim 13, characterized by a Device for applying two sets of coding gradient pulses during the first and second basic scan sequences to achieve a three-dimensional image display. 19. The device according to claim 18, characterized in that the device for placing two sets of devices to give up a first sentence of the two sentences of coding gradient pulses along the axis of the first magnetic gradient pulse and the second set of coding gradient pulses along the axis of the second has magnetic gradient pulse.   20. Device according to claim 13, characterized by a Device for applying the spoiler and phase shift pulse in the basic sequences. 21. Device according to claim 13, characterized by a Device for applying counterfeit gradient pulses between the first and second RF pulses. 22. Device according to claim 13, characterized by a Device for applying asymmetrical gradient pulses to eliminate unwanted signals. 23. Device according to claim 13, characterized by a device for applying n first RF pulses before the second RF pulse is applied, each of the n first pulses being a different offset frequency for obtaining signals from selected volumes of interest. 24. Device according to claim 13, characterized by a device for giving up the second basic scanning sequence, to get signals from a selected strip achieve.